Papers with TaDSE

1 papers
Template-assisted Contrastive Learning of Task-oriented Dialogue Sentence Embeddings (2026.acl-long)

Copied to clipboard

Challenge: Annotating and gathering utterance relationships in dialogues is difficult, while token-level annotations, entities, slots and templates, are much easier to obtain.
Approach: They propose a template-aware augmentation method that utilizes template information to learn utterance embeddings via self-supervised contrastive learning framework.
Outcome: The proposed method improves on five benchmark dialogue datasets and shows that it is more efficient than previous SOTA methods.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations